This post has been a long time coming. Originally it was going to be an entire series of posts (remember when I used to post more than 5 items a month? Good times), going into an inane amount of detail, and featuring case studies on every single aspect of the featured subject.

It is about doors.

You might be thinking that I couldn’t possibly be dull enough to fill post upon post with doors. The idea was going to be to examine what is arguably the most common object in everyday human interaction and go into why it is, pardon my French, fucked up so often. I would start with looking at each component (primarily the handle, but any windows play an important part in interaction) and show how they are usually done so badly as to make the primary task of opening a door more difficult for people.

It was to be an opus on how, by paying attention to the task at hand and examining small details, to best design an object. But all you really need to know is this:

Excluding life threatening situations, door design is the single most common bit of bad design. Remember: handles mean pull, panels mean push. Never change this. For large values of never.

Ever have to design a door? Now you know how. Ever need to design a car, pen, website or anything, bear the above lesson in mind.

Yes, it’s late but, in my defence, we’ve been having remarkably good weather these last few weeks. This month there are six films up for review. Onwards.

First up is the suprising Silent Hill; surprising because despite some fairly impressive special effects with which to play (the melting between worlds), the film is incredibly bad. A bad mix of the first two Silent Hill games, the plot doesn’t really know where it is going. The acting is as b-movie as you would expect, with Sean Bean putting in a terrible performance (the man cannot do American accents, don’t make him try). Couple this with a plot unravelled in a haphazard mixture of dream sequences, narration, flash backs and every other bad ploy taken by poor film makers in recent years and you have a film which does not do the plodding source material justice.

Slither is an excellent comedy horror film that finds the right balance between laughs, characters and a hokey plot. Sure, it’s a fairly textbook premise (aliens body snatchers 101), but the spirit of the film is good natured and never tries to be more than that. It won’t be winning any awards, but worth the effort.

On the action front, Mission:Impossible 3 is exactly that: a dumb action film. Arguably better in many ways that it’s fairly shoddy predecessors, this take is handled with more style but the same logical gaps that make the series exasperatingly painful. Hoffman did the best with what he had to work with and Cruise… well, Cruise runs about a bit. Expect to see the same old rope tricks and face changes.

An American Haunting is a schizoid horror film. Set primarily in the 1700s, it chops and changes between wanting to be a ghost story, to something far more pedestrian but chilling, to a ghost revenge story, all bookmarked by a pointless narration in modern times. The desperation of the family comes across as laboured, and the decline of the main characters as expected. Terrible.

Scott Ryan has put together a delightfully offbeat mockumentary in the form of The Magician, the story of a hired killer who is eventually offered a large cash sum not to kill a victim. The subject matter is dark and the humour is too, coming mostly from the bizarre conversations between the central character, has cameraman and would-be victim. Funny, different, good: go see.

The gross-out comedy stuffing is pretty stale these days, but Waiting… is a decent example of serving it well. The guts of this comedy revolve around a simple recurring joke involving male nudity (if you haven’t been introduced to “The Game” then it’s not my place to get you in), but it hits just the right spot. The character development etc are as weak as one might expect but like all films of this genre, it’s about the quirks of the characters as they stand. Not a masterpiece by any standards, but still a fairly stand-up, dumb comedy.

Finally, Down In The Valley stars Ed Norton as a cowboy who finds himself a cowgirl in the sprawl of Los Angeles. Things, inevitably, go wrong and Ed goes a bit off the wall. While Norton puts in a decent performance, the film itself is fairly meandering and badly cut. Dull, turgid, bad.

The winner for May is easily The Magician.

Linux and filesystem experts, I need your help. A while back, I started looking at installing Ubuntu after discussing various options with Stuart Langridge. I’ve subsequently burned 3 different releases of said OS to DVD, up to and including the latest 6.06 LTS x64 version. Playing with the live CD on a few occassions has convinced me that it is the right idea: it’s a tidy little Linux distro that gets out of your way until you want to really start messing around with it. I have not, however, installed it yet for two reasons. Primarily, it has taken me a fair amount of time to investigate everything I needed to know about Ubuntu before being happy with it, making sure I’ve thought through any potential problems. I’m glad I did that because I’ve hit a real problem, which brings me to my second reason.

My computer is an AMD64x2 based machine with two 200Gb SATA drives, both of which were formatted as NTFS volumes. The first of those is the boot drive, which contains installed programs and games and has almost 125Gb free. The secondary drive is my data drive, containing video, music, pictures, websites and a whole load of other application data I don’t want to lose. It is full, give or take a few gigs.

The problem is the data drive. I want a volume that I can read and write from either windows or linux (Mac support would be nice, but not essential) but, as the drive is NTFS, the linux support is poor for writing and this is data I don’t want to lose. I’ve been giving it some thought and have a few ideas:

1. Long, bitty migration. I could offload a good chunk of the data drive to the non-data drive, create a new partition on the data drive (in the new freespace) with a better filesystem (whiche one), migrate the data back to that partition, and repeat with the other section. This strikes me as potentially messy.
2. The Linux-NTFS project. Although they now have NTFS support, it’s at the expense of “speed and stability”. I don’t know if that’s a worthwhile trade-off.
3. Checking out NTFS for linux. I’d be willing to pay for the software if it safely met my needs, but I am not sure it will.

Of course, those options all miss a key point: the data drive is absolutely full. I will need more space at some point though the largely free non-data drive can be used for now, so this might be a perfect opportunity. I think (though I am not sure) my motherboard, an Asus A8N32-SLI Deluxe, only has two SATA slots and that makes just shoving in another disk tricky. So I’m also looking towards NAS solutions:

1. Buying a NAS box from Infrant. It’s quite a cost up front to get even a diskless NAS box, but it’s the sort of investment that might be worth it. While I’m wary that nothing will be compatible forever, an ethernet-accessed blackbox with upgradeable firmware seems like a safe bet for at least the medium term. Cost might be an issue, but I’m also not sure about all the access issues. Presumably it uses some standard access protocol (Samba-based most likely), but it’s been a while since I have looked at the issues around network storage.
2. Build my own NAS box based on FreeNAS. This would be a fun little project, but I’m not sure it would be any cheaper than buying a more sophisticated box given a) the cost of components, b) time taken, c) reliability, d) feature set, and e) lifespan.

So, gentle readers, have I missed an obvious or better option? Do these options look good? Which is best? Are they problems I have overlooked? All comments and suggestions welcome.

I’ve previously written about open source project home pages and thought it was about time to look at another one, namely the Banshee Project: a Linux-based music player and organiser which takes heavy inspiration (and features) from iTunes.

Most aspects of the site are spot on. The design is clean and friendly, the content is well organised and the copy used helpful and concise. Best example is the opening line from the front page:

With Banshee you can easily import, manage, and play selections from your music collection.

That’s how you sell a product: a clean, and simple explanation of the benefits it aims to provide. After a few more sentences to explain additional functionality, it allows new users to look at more information via a “Read more” link, and continues with download details, bug channels, mailing lists and other details more experienced users are looking for. That’s near impeccable flow in design.

The main thing the site gets wrong is the download links: they are below the page fold and/or poorly labelled. “View this release” and “Getting Started” are not what people are looking for; they want a link that says “Download”. Put that in somewhere above the fold, and make the download experience as painless as possible (one-click to get the executable) and you’ve got a marked improvement. Admittedly this is more difficult for a Linux application than a Windows one, given differences in distros, but it should be easier than it currently is.

All-in-all, a good example of an open-source project that is getting it (mostly) right.

I find it interesting how much more obsessive a small group of people (designers and technology types) are about small aesthetic details in interactive devices than the vast majority of people who use them.

For example, give most people a computer keyboard and they will be content to just use it, regardless of whether it’s a bog standard model or a bizarrely contoured model. Those few others, though, can tell if the keyboard is any good by putting their hands on it for a few seconds. They’ve been using computers long enough to know what a keyboard should feel like from the layout of the board, to the size of the keys, to the relative heights of keys. Lots of little details that just click into place. I can’t tell you the number of discussions I’ve had with various lunatics about whether clicky keys (think an old-school Spectrum keyboard) are best or gummy silent keys are superior (the answer is clearly neither, you want quiet keys that have a good responsive action). I don’t think anyone can explain the horror felt by one of these people when asked to use a mouse that is too small, doesn’t have a wheel (I personally CANNOT do it), or only has one button (Mac users of yesteryear, you are wrong).

Is this ridiculous compulsive behaviour a bad thing? No. We represent the top end of the market, everything is shaped by us. It’s because we care about the details that no-one else really has to.